Pressure Factors: How Temperature, Powder, and Primer Affect Pressure Denton Bramwell [email protected] The Curiosity Cycle Initiated…Again It’s well accepted that temperature influences pressure. But the temperature of what? Is it the ambient temperature? Or the temperature of the barrel? Or the temperature of the powderi? Or is it some combination of all these factors? If so, do different powders and primers respond differently to temperature? Is there a combination of reloading components that works better than the others? Soon after I instrumented two rifles with strain gages, and started making pressure measurements, I became convinced that barrel temperature influenced pressure. I started taping a thermocouple to the barrel, just forward of the receiver, and doing all my pressure measurements at close to constant barrel temperature. Is that thermocouple really necessary, or is it excess baggage? With my trusty PressureTraceii in hand, I decided to find out. The Experiment Designed The most common way to do experiments is to hold all factors constant except one, and to look at the influence that the one variable factor has on the outcome. The usual statistical tool for evaluating the results is the Student’s T Test. The T Test is an excellent test, but this is a very inefficient approach. The really efficient experimental tools let you look at multiple variables, and all of their interactions, at the same time. You can investigate five factors and all 26 of the possible two, three, four, and five-way interactions for the same amount of work as a T Test, and with the same “clarity” as a T Test. The simple T Test just investigates one variable and no interactions iii. The efficient way is to gather data in a special way, known as a “balanced design”. This is a simple example of a balanced design, where two states are allowed for each input variable, A, B, and C. For each combination of A, B, and C, the output variable, such as peak pressure, is recorded. A B C Output Variable Low Low Low High Low Low Low High Low High High Low Low Low High High Low High Low High High High High High One complete set of data like this is called a “replicate”. More than one replicate is required for a lot of the analysis that is usually needediv. 1 The math “magic” is that we can evaluate each variable, and each interaction, as though the others were not there. If you look at the states of B and C that happen with the low state of A, you will find that you have equal and opposite pairs of both B and C. The same is true of each state of each variable and each interaction. That’s why the balanced design is so useful. This is a very efficient scheme for investigating systems with multiple input variables that might interact, like rifle chamber pressure. I chose four input variables. 1. Type of powder (IMR4350 vs H4350)v, 2. Type of primer (CCI 200 large rifle primers vs. CCI 250 magnum large rifle powders), 3. Barrel temperature (50 vs. 95 degrees F, and 4. Powder temperature, (45 F vs. 100 F). It’s also easy to investigate more than one output variable (result) at a time, and since the PressureTrace gives both pressure and risetime, I took data for both output variables. I wanted to investigate ambient temperature as the fifth variable, but it is just too awkward to try to keep the barrel and ammunition at pre-selected temperatures, independent of ambient temperature. I gave up on that one. I don’t think ambient temperature is important, anyway, except for its effect on powder temperature and barrel temperature. I would not be too hasty to apply these results to other cartridges, powders or primers. They may behave differently. Also, there is no practical way to be sure that the particular jars of powder or boxes of primers I have on my shelf are representative of all lots of IMR or Hodgdon 4350 powder, or CCI primers. The rifle used for the experiment was a Finnish M39 milsurp. It does run a little less pressure than modern firearms do, but it is a fairly heavy rifle, which is pleasant when you have a lot of shooting to do, and it was already instrumented and ready to go. I wondered if I could really separate barrel temperature from powder temperature. As soon as you put a cold cartridge in a hot barrel, it starts to warm. The outer layers warm first, and most of the volume is in the outer layers. Even if you shoot single-shot style, and fire as quickly as you can, will the heat of the barrel warm a cold cartridge before you can fire? I prepared a special cartridge with bullet, casing, and powder, but no primer, with a thermocouple in the center of the cartridge. I packed the cartridge in a plastic bag full of ice, and watched the temperature at the core start to fall, due to roughly a 50 F temperature difference. The straight-line portion of the curve fell at about 1/4 degree F per second, which leads me to believe that if you load and shoot quickly, barrel temperature will have only a small effect on powder temperature. You can separate the two temperature variables, if you’re quick. 2 7.62x54R casing, filled with very old H4831 powder, and with thermocouple inserted into the center of the case, through the primer hole. 3 This also indicates that the SAAMI practice of letting ammunition sit at test temperature for 24 hours before doing a test is overly cautious. 24 hours isn’t going to hurt anything, but an hour or two is actually ample. To the Range! Early in September, with the dog days of August behind us, and a pleasant 81 F ambient temperature, I headed out to the range to do the “hot barrel” part of the test. I strapped a thermocouple to the barrel, and fired several shots to bring the barrel up to 92 F. For the remainder of the test, I kept the barrel between 92 and 98 F. Ammunition was held at my low (45 degrees F) and high (99 degrees F) temperatures, and fired as quickly as I could get a cartridge out of storage, load, and fire it. The whole cycle took less than 10 seconds. Test cartridges packed for the test. A few pounds of river rock in the bottom of the cooler regulated the temperature. A towel, folded on top of the ammunition boxes, provided additional insulation. The cold weather testing turned out to be a bit more challenging. December did provide exactly the day that I had been waiting for. It was a Saturday, so the outdoor range was open. The ambient was about 40 F, which is cold enough to let me keep the barrel close to 50 F. I let the cold ammo sit overnight in an unheated shop, which brought it very close to 45 F. I put power resistors under the river rock of the hot ammo cooler, and applied power, to bring it up a bit above 100 F. I even took a husky battery to the range with me, so I could warm the ammo anytime the thermocouple in the river rock said it was cooling off too much. Unfortunately, after about six rounds, my thermocouple meter abruptly started giving strange readings. Apparently, the cold had 4 gotten to circuitry. With no way to regulate barrel temperature, there was nothing to do but to pack up and go home. Since I already had a three-factor balanced design completed with a warm barrel, I decided to simply pursue the effect of barrel temperature on pressure with a separate, single factor regression test at an indoor range. In that case, everything was held constant except the barrel temperature, which was allowed to rise as shots were fired at about one minute intervals. This will not allow discovery of barrel temperature interactions, but it is a satisfactory solution. The loads fired were 150 grain bullets, propelled by Varget. Peak pressure and pressure risetime were measured as a function of barrel temperaturevi. Test Results and Conclusions Main Effects Plot (data means) for Peak Pressure n do dg R m Ho IM lar nu 50 50 gu ag 43 43 re m 45 99 50800 I S 49800 P , e r u s 48800 s e r P k 47800 a e P 46800 powder type primer type TCartridge, F 1. This is the Main Effect plot for pressure vs. powder type (Hodgdon and IMR 4350), type of primer (CCI200 and CCI250), and the temperature of the cartridge (45 and 99 F). Going from Hodgdon H4350 to IMR 4350 reduces pressure by 3624 PSI. There is about one chance in 1,000 of getting a difference as big as this one, just from the random chance in the experiment. This effect is real. 5 Going from the CCI200 (regular primer) to the CCI250 (magnum primer) produces 432 PSI additional peak pressure. There is a 63% chance of getting a difference this large, just from random variation, so this difference is attributed to random variation. That is, switching from CCI regular large rifle primers to CCI magnum large rifle primers, produced no detectable effect on pressure. It’s worth noting that in a separate test, I found that switching to a magnum primer actually reduced peak pressure slightly, with 4831 powder in this cartridge. Going from a powder temperature of 45 F to a powder temperature of 99 F increases pressure by 3797 PSI, or about 70.3 PSI per degree F.